In recent years, a study of an ad hoc network has been progressed in which a plurality of communication apparatuses each referred to as a node are connected with each other in an autonomous distributed manner. In the ad hoc network, a time synchronization source that is a node holding a reference time used for time synchronization between nodes is provided, and times are synchronized among nodes using the time synchronization source as a center.
The time synchronization source is, for example, a node that receives a time from a time calibration server (NTP server) provided on the network, or a node that is directly connected with a handy terminal (HT) used for the meter-reading of data and sets a time of the present node. When a node that is directly connected with the HT serves as a time synchronization source, the node acquires and holds the time of the HT as a reference time. Furthermore, each node on the ad hoc network sets the time of the present node based on a reference time included in time information propagated sequentially from a node that is the time synchronization source. In this manner, in a network formed in conjunction with ad hoc communication, a node group including a plurality of nodes synchronized with one node that is the time synchronization source for each network formed is formed. That is, in the network formed in conjunction with the ad hoc communication, an independent node group is formed with respect to a node that is the independent time synchronization source.
Here, in the network formed in conjunction with the ad hoc communication, when a plurality of ad hoc networks are brought into contact with each other, communication is performed across the boundary of each ad hoc network. However, when nodes belonging to the respective ad hoc networks have the respective times different from each other, the nodes are incapable of communicating with each other. For example, to consider a case where each of a plurality of ad hoc networks has a node that is a time synchronization source based on a time acquired from an HT, when a time acquired from an HT and a time acquired from the other HT are different from each other, reference times of the nodes each of which is a time synchronization source that acquires a time from an HT are different from each other and hence, node groups connected with different time synchronization sources are incapable of communicating with each other. Therefore, in order for a plurality of ad hoc networks to communicate with each other, synchronizing times among a plurality of node groups is performed.
As a technique that synchronizes in terms of times among a plurality of node groups connected with different time synchronization sources, there has been proposed a technique that sets priorities to the node groups, and synchronizes a node positioned at the boundary among the node groups in terms of a time with the highest priority node group (see Japanese Patent No. 3865223). In this technique, a node synchronized in terms of a time with the other highest priority node group also transmits the time after synchronization to a node adjacent thereto in the node group thereof. Accordingly, synchronization in terms of times among a plurality of node groups can be achieved with the highest priority node group.
In the above-mentioned conventional technique, a node is synchronized in terms of a time with one of the time synchronization sources and hence, times among nodes with the identical time synchronization source can be synchronized with each other and the nodes can communicate with each other. However, there arises a difference in time between a node that changes the time synchronization source thereof and each node belonging to the node group synchronized in terms of a time with a time synchronization source before the change depending on the difference in time between the time synchronization sources and hence, there has been a drawback that the communication with a node previously capable of communicating becomes impossible by changing the time synchronization source.